More satellite stuff, including explanations, a bit

When I wrote my prev post (BTW, the same caveats about data handling that I made there still apply), I was in a bit of a hurry and included little in the way of background. So... there is some good stuff at http://en.wikipedia.org/wiki/Satellite_temperature_measurements (well I wrote most of it). That page also includes links to various groups, but RSS is Remote Sensing Systems and S+C/UAH is here and other places; but their pages are in a bit of flux because they haven't caught up with their latest correction yet! The RSS pages have some nice descriptions of how the stuff all works.

Anyway, I've done a few more pics: here are maps of the trends (1979-2004, from annual mean data) from S+C vn5.2, 5.1, RSS and 5.2-5.1:

The last is quite startling, to me: it looks like an essentially zonally symmetric correction in the tropics. Odd.

And then the zonal means of the trends:

Also interesting (perhaps) is the trends to-year for the two series:

year 5.2 5.1
1992 0.0009 -0.0005
1993 -0.0025 -0.0046
1994 -0.0017 -0.0045
1995 0.0015 -0.0014
1996 0.0021 -0.0008
1997 0.0028 -0.0001
1998 0.0099 0.0070
1999 0.0089 0.0058
2000 0.0080 0.0046
2001 0.0090 0.0054
2002 0.0108 0.0072
2003 0.0117 0.0080

These are the trends calc from the monthly data, from jan 1979 to dec of (year). They aren't quite the same as the ones on the wiki page (for 5.1) but thats probably down to minor processing (my area avg vs theirs, perhaps). So on my version of their v5.2, the negative trends stop in 1994, 3 years earlier than for 5.1.

Meanwhile... over at Deltoid, Scott Church commented: Note that these are TLT trends, so they won’t compare directly with current versions of RSS MSU trends which are for the middle troposphere (the corresponding UAH trends for an RSS or Prabhakara comparison will be TMT). I'm actually a teensy bit confused about the various products and levels. The S+C vn5.1/5.2's are definitely comparable. The RSS data may not quite be. The RSS stuff is their TMT product, which is a merge of MSU channel 2 and some AMSU data.


Anonymous said...

Although I don't exhibit other obvious signs of masochism, I periodically visit Tech Central Station to see what the Dark Side is currently spouting. Just now I found this interesting squib from Roy Spencer: http://www.techcentralstation.org/062005G.html posted on May 20th. Given that it must have been written around the same time he became aware of the need for the satellite revision, one has to wonder what he was thinking.

Scott Church said...

The main differences between UAH (Spencer and Christy) TLT and the TMT satellite records are as follows (I'll try to be brief!).

The MSU devices (Microwave Sounding Units) on these satellites take their temperature measurements using repeated scans of downward looking sensors. These sweep from left-looking to right-looking views across the satellite's orbital path, passing through straight down (nadir) along the way. They have multiple channels that sense at different frequencies, each being most sensitive to a particular layer of the atmosphere. The one that's of interest here is Channel 2, which gets its signal from a layer of the atmosphere that peaks at around 800 hPa and dwindles out above 50. Channel 2 measurrements are a good representation of "middle troposphere" temperature--that is, the "bulk" temperature of the troposphere below the tropopause but above the surface. For a clearer picture see the paper I wrote on the satellite record at my web site. In particular, note Figure 7.

The problem is, what's of interest to everyone is the lower troposphere (e.g. - 1000-800 hPa roughly) and this isn't directly measured by MSU's (it needs to be remembered that these devices, which have been in service since the 70's, were not designed with global warming data gathering in mind). About 15 years ago Spencer and Christy discovered that the side-looking and straight-down views of the MSU sensors across each of its sweeps could be factored and combined in a way that would give a temperature for the lower part of the troposphere. Because it is derived from the MSU Channel 2 data by formula, this was referred to as a "synthetic" channel and designated "2LT". With the latest versions, UAH has begun designating this as "TLT".

So TLT is UAH's lower troposphere temperature product that is derived from the actual MSU Channel 2 measurements. What they designate as "TMT" is their uncorrected Channel 2 product that gives the "middle" troposphere.

The contention over all this due partly to the fact that this TLT derivation process applied to side looking sensor views has a fair amount of sampling noise. The RSS and Prabhakara teams avoided this by simply sticking with the middle troposphere and uncorrected downward looking views. The Fu et al. method applied to Channel 2 measurements corrects these to yield something very similar to the band TLT represents and would provide a more direct TLT comparison. So in summary,

Lower Troposphere: TLT -> RSS plus Fu method correction.

Middle Troposphere: TMT -> RSS.

Hope all that helps!

Brad Tittle said...

When I was getting my Engineering Degree, my papers were routinely marked up in red anytime I did any "zooming" of data to make it look pretty.

I look at all of the climate "anomalies" graphs and wonder if any of the science professors ever get their red pens out. Showing "anomalies" might be reasonable, but only if the raw data is also displayed. Personally I would prefer to see the graph in deg K starting at 0 K (or R if you prefer), but such transparency seems to be lost of climate researchers seeking more funds. The worrisome aspects of climate change become rather difficult to see when put in perspective.

Of course this is why my Professors kept marking up my work in this particular area. Maintain perspective so you make good decisions.

Every time I see an Anomalies graph published without a contextual Temperature graph next to it, my BS meter gets pegged.

This has been created by the Excel Generation of scientists (AKA Spreadsheet Man). I hope that rationality returns to the sciences eventually, but I am not hopeful with the nonsense that abounds today.